Talk ABRF 2015 (Gunnar Rätsch)

Analysis Tools for RNA-seq and
Isoform Characterization
Slides: t.co/nc7siKRm6W
Gunnar R¨atsch
Biomedical Data Science Group
Computational Biology Center
Memorial Sloan Kettering Cancer Center
gxr #RNA #MMR #SplAdder #riboDiﬀ #Cancer

Biomedical Data Sciences Group
Facts
Cost of collecting data drops, amounts increase exponentially.
We have more data than accurate algorithms.
Group’s research
Data Science Algorithms, Models & Tools
Machine Learning,
Bioinformatics.
Biology & Medicine Problem Setting & Goals
RNA processing regulation,
Clinical data analysis.
c Gunnar R¨atsch (cBio@MSKCC) Tools for RNA-seq and Isoform Characterization ABRF Annual Meeting 2015, St. Louis 1
Memorial Sloan-Kettering Cancer Center

Learning About the Central Dogma
Goal: Learn to predict what these processes accomplish:
Given the DNA, . . . , predict all gene products
f (DNA, ) = RNA g(RNA, ) = protein
Estimating f , g amounts to cracking the codes of
transcription, epigenetics, splicing, . . .

RNA-seq based Transcriptome Characterization

RNA-seq based Transcriptome Characterization
oqtans.org
cloud.oqtans.org

Transcript Quantitation and Dependence on Alignments
0 1 2 3 4 5 6 7 8
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
maximal number of mismatches
Pearsoncorrelationwithtrueabundance
True read origins
Alignments to
genome
92%
53%
84%
False alignments, multi-mappers etc. lead to weaker results
Simulated human reads from transcripts of known abundance (Fluxsimulator, [Sammeth, 2009]), 3% error rate, alignment w/
PALMapper [Jean et al., 2010], quantiﬁcation w/ rQuant [Bohnert et al., 2009], Person correlation over considered transcripts.

bioRxiv dx.doi.org/10.1101/017103
Efficient BAM file postprocessor for RNA- & DNA-seq
100M alignments in 20 minutes (10 threads)
Suitable for large-scale projects
Improved accuracy for transcript quantification and prediction
Open Source bioweb.me/mmr (C++)

Multiple Mapper Resolution
Principle (Iterated over all reads, N times)
Use the change of local coverage around read mapping ...
... and use its smoothness to identify “better” mapping location
location 1 location 2
Coverage

Read not mapped to location 1 ... ... but mapped to location 2
Read pairCoverage

+
Read pair Variance measure Evaluation windowCoverage Average coverage

+
+
Read pair Variance measure Evaluation windowCoverage Average coverage
Read mapped to location 1 ... ... and not mapped to location 2
Assignment1Assignment2

Results for simulated DNA-seq
Smooths coverage as expected on an artiﬁcial dataset
Simulated reads from tiling a part of A. thaliana genome, alignment w/ PALMapper [Jean et al., 2010] (with -a option), visual-
ization with IGV [Robinson et al., 2011].

Results for simulated DNA-seq
Smooths coverage as expected on an artiﬁcial dataset
≥ ≥
Simulated reads from tiling a part of A. thaliana genome, alignment w/ PALMapper [Jean et al., 2010] (with -a option), visual-
ization with IGV [Robinson et al., 2011].

Results for simulated RNA-seq
Improves performance of transcript quantiﬁcation
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
1
0 1 2 4 6 all
Correlation
Mismatches
original alignments
best alignment only
MMR treated alignment
Simulated reads (75nt) from subset of human annotated transcripts with Fluxsimulator [Sammeth, 2009], PALMapper alignments
[Jean et al., 2010], rQuant quantitation Bohnert et al. [2009], Pearson correlation over all considered transcripts.

LARGE-SCALE BIOLOGY ARTICLE
Nonsense-Mediated Decay of Alternative Precursor mRNA
Splicing Variants Is a Major Determinant of the Arabidopsis
Steady State TranscriptomeC W
Gabriele Drechsel,a,1 André Kahles,b,1 Anil K. Kesarwani,a Eva Stauffer,a,2 Jonas Behr,b Philipp Drewe,b
Gunnar Rätsch,b and Andreas Wachtera,3
a Center for Plant Molecular Biology, University of Tübingen, 72076 Tuebingen, Germany
b Computational Biology Center, Sloan-Kettering Institute, New York, New York 10065
ORCID IDs: 0000-0002-3411-0692 (A.K.); 0000-0001-5486-8532 (G.R.); 0000-0002-3132-5161 (A.W.).
The nonsense-mediated decay (NMD) surveillance pathway can recognize erroneous transcripts and physiological mRNA
such as precursor mRNA alternative splicing (AS) variants. Currently, information on the global extent of coupled AS and NM
The Plant Cell, Vol. 25: 3726–3742, October 2013, www.plantcell.org ã 2013 American Society of Plant Biologists. All rights reserved.

Analysis of alternative isoforms with RNA-seq data
Analyses known and identifies novel splicing events
Quantifies & visualizes splicing-related data
Suitable for large-scale projects (1000’s of samples)
Improved accuracy for transcript quantification and prediction
Open Source bioweb.me/spladder (python)

SplAdder Ideas
Major Problems in Transcriptome Analysis
1 Gene annotations are incomplete and often inaccurate
2 Whole transcript isoforms are diﬃcult to predict/quantify

SplAdder Ideas
Solution
Augment annotation with RNA-Seq evidence
Use single splicing events instead of full transcripts

SplAdder Ideas
Solution
Augment annotation with RNA-Seq evidence
Use single splicing events instead of full transcripts
Annotation
Alignment
Data
Augmented
Splicing
Graph
Detected
Splice
Events
Quantified
Splice
Events
Differential
Analysis/
sQTL Tests
Kahles et al., bioRxiv, 2015

SplAdder Graph Augmentation
Principle
Collapse annotated transcripts into graph representation
Use RNA-Seq evidence to add new nodes and edges
T1E1 T1E2 T1E3 T1E4
T2E1 T2E2 T2E3
T3E1 T3E2 T3E3
T4E1 T4E2

Principle
T1E1 T1E2 T1E3 T1E4
T2E1 T2E2 T2E3
T3E1 T3E2 T3E3
E1 E3 E4 E6
E2 E5
T4E1 T4E2

Principle
coverage
split alignments
New cassette exon

Principle
coverage
split alignments
New cassette exon
coverage
New retained intron

coverage
split alignments
A New cassette exon
coverage
B New retained intron
split alignments
C New intron
split alignments
D Alternative splice sites on both intron ends
split alignments
E New start-terminal node / New end-terminal node
split alignments
F Alternative 3’splice site / New end-terminal node
split alignments
G Alternative 5’splice site / New start terminal node
split alignments
H New exon skip

SplAdder Event Extraction
E1 E3 E4 E6
E2 E5

SplAdder Event Extraction
E1 E3 E4 E6
E2 E5
E1 E3 E4 E6
E2 E5
E1 E3 E4 E6
E2 E5
E1 E3 E4 E6
E2 E5
E1 E3 E4 E6
E2 E5
Exon Skip
Multiple Exon Skip
Alternative 5’Splice Site
Intron Retention
Alternative 3’Splice Site

SplAdder Event Quantiﬁcation and Visualization
Exon Skip

Exon Skip
a
cb

Exon Skip
a
cb
PSI =
b + c
2 · a + b + c

Alternative 5’ Site
a
b
Exon Skip
a
cb
PSI =
b + c
2 · a + b + c
PSI =
b
a + b

Summary
SplAdder eﬀectively augments the annotation
Enables quantitative analysis of events instead of transcripts

Splicing Analysis Across Multiple Cancer Types
Goals
1 Identify cancer-speciﬁc splicing patterns
2 Identify variants regulating splicing in same gene (cis)
3 Identify variants regulating splicing in other cancer genes (trans)
TCGA provides RNA-seq and matching exome data
RNA-seq Find & quantify splicing events
Exome Identify variants in exons & ﬂanking intronic regions

Splicing Variation Across 4,700 Samples
Event Statistics Exon Skip Eve
A B
Analysis of a total of 4,700 RNA-seq samples from TCGA normal (tn), TCGA tumors (tc), Encode (ec) and Geuvadis (gv). Align-
ment w/ STAR [Dobin et al., 2013], analysis w/ SplAdder (SplA) and Gencode annotation (Anno). Figure from [Kahles, 2014].

Uniform analysis of Large-Scale RNA-seq Data
Large-scale Compute
4,700 RNA-seq libraries (≈100 TB)
⇒ STAR ≈ 6 CPU years
⇒ SplAdder ≈ 0.5 CPU years
[Kahles et al.]
Uniﬁed community resources
Docker with ICGC RNA-seq alignment SOP
bioweb.me/ICGC-RNA-SOP
Syncronize with Encode, gTex, TCGA, . . .
[ICGC PCAWG-3 WG]

Analysis of Ribosome profiling and RNA-seq data
Study translation efficiency
Adjusts for expression differences
Accurate method based on dispersion estimates and GLMs
Open Source bioweb.me/ribodiff (python)

Application to Ribosome Profiling
Found a motif that was strongly in enriched in detected genes:
G-quadruplex structures
Application to Ribosome Profiling
Found a motif that was strongly in enriched in detected genes:
G-quadruplex structures
(Analysis based on related but different strategy [Wolfe et al., 2014].)

Summary
MMR improves alignment choice for multi-mappers
⇒ Helps improving accuracy of tools like Cufflinks
SplAdder identifies, quantifies & visualizes alternative splicing
⇒ Finds unannotated alternative splicing, tumor/normal splicing
differences; splicing reprogramming; sQTLs
riboDiff accurately detects differential translation efficiency
⇒ Ribosome footprinting revealed RNA G-Quadruplex elements in
5’ UTR that interacts with compound via eIF4a
Tools (+ six other ones) are open source and available
⇒ ratschlab.org/tools
⇒ (more) Docker images come soon . . .

Acknowledgements
R¨atsch Laboratory
Andre Kahles
Yi Zhong
Philipp Drewe @ MDC Berlin
Theofanis Karaletsos
Kjong Van Lehmann
Jonas Behr @ ETH Basel
Regina Bohnert @ Molecular Health
Geraldine Jean @ University of Nantes
Cancer Biology
Guido Wendel
Kamini Singh, . . .
Cancer Genomics Projects
Angela Brooks, Broad
Alvis Brazma, EBI
Matt Wilkerson, UNC
Niki Schultz, MSKCC
Chris Sander, MSKCC
Funding from MSKCC, Max Planck Society,
European Union, German Research Foundation,
Geoﬀrey Beene Foundation & NIH
Thank you!

References I
J. Behr, G. Schweikert, J. Cao, F. De Bona, G. Zeller, S. Laubinger, S. Ossowski,
K. Schneeberger, D. Weigel, and G. Rätsch. Rna-seq and tiling arrays for improved gene
finding. Oral presentation at the CSHL Genome Informatics Meeting, September 2008. URL
http:
//www.fml.tuebingen.mpg.de/raetsch/lectures/RaetschGenomeInformatics08.pdf.
R. Bohnert, J. Behr, and G Rätsch. Transcript quantification with RNA-Seq data. BMC
Bioinformatics, 10(S13):P5, 2009. URL
http://www.biomedcentral.com/1471-2105/10/S13/P5.
RM Clark, G Schweikert, C Toomajian, S Ossowski, G Zeller, P Shinn, N Warthmann, TT Hu,
G Fu, DA Hinds, H Chen, KA Frazer, DH Huson, B Schölkopf, M Nordborg, G Rätsch,
JR Ecker, and D Weigel. Common sequence polymorphisms shaping genetic diversity in
arabidopsis thaliana. Science, 317(5836):338–342, 2007. ISSN 1095-9203 (Electronic). doi:
10.1126/science.1138632.
Alexander Dobin, Carrie a Davis, Felix Schlesinger, Jorg Drenkow, Chris Zaleski, Sonali Jha,
Philippe Batut, Mark Chaisson, and Thomas R Gingeras. STAR: ultrafast universal RNA-seq
aligner. Bioinformatics (Oxford, England), 29(1):15–21, January 2013. ISSN 1367-4811. doi:
10.1093/bioinformatics/bts635. URL http://www.ncbi.nlm.nih.gov/pubmed/23104886.
Mitchell Guttman, Manuel Garber, Joshua Z Levin, Julie Donaghey, James Robinson, Xian
Adiconis, Lin Fan, Magdalena J Koziol, Andreas Gnirke, Chad Nusbaum, John L Rinn,
Eric S Lander, and Aviv Regev. Ab initio reconstruction of cell type-specific transcriptomes
in mouse reveals the conserved multi-exonic structure of lincrnas. Nat Biotechnol, 28(5):
503–10, May 2010. doi: 10.1038/nbt.1633.

References II
G Jean, A Kahles, VT Sreedharan, F De Bona, and G Rätsch. Rna-seq read alignments with
palmapper. Curr Protoc Bioinformatics, Unit 11.6, 2010.
Andre Kahles. Novel Methods for the Computational Analysis of RNA-Seq Data with
Applications to Alternative Splicing. PhD thesis, University of Tübingen, Tübingen,
Germany, September 2014.
G. Rätsch and S. Sonnenburg. Accurate splice site detection for Caenorhabditis elegans. In
K. Tsuda B. Schoelkopf and J.-P. Vert, editors, Kernel Methods in Computational Biology.
MIT Press, 2004.
G. Rätsch, S. Sonnenburg, and B. Schölkopf. RASE: recognition of alternatively spliced exons
in C. elegans. Bioinformatics, 21(Suppl. 1):i369–i377, June 2005.
James T. Robinson, Helga Thorvaldsdóttir, Wendy Winckler, Mitchell Guttman, Eric S. Lander,
Gad Getz, and Jill P. Mesirov. Integrative genomics viewer. Nature Biotechnology, 29:
24–26, 2011.
M. Sammeth. The Flux Simulator. Website, 2009. http://flux.sammeth.net/simulator.html.
Gabriele Schweikert, Alexander Zien, Georg Zeller, Jonas Behr, Christoph Dieterich,
Cheng Soon Ong, Petra Philips, Fabio De Bona, Lisa Hartmann, Anja Bohlen, Nina Krüger,
Sören Sonnenburg, and Gunnar Rätsch. mgene: Accurate svm-based gene finding with an
application to nematode genomes. Genome Research, 2009. URL
http://genome.cshlp.org/content/early/2009/06/29/gr.090597.108.full.pdf+html.
Advance access June 29, 2009.

References III
S. Sonnenburg, G. Rätsch, A. Jagota, and K.-R. Müller. New methods for splice-site
recognition. In Proc. International Conference on Artificial Neural Networks, 2002.
Sören Sonnenburg, Alexander Zien, and Gunnar Rätsch. ARTS: Accurate Recognition of
Transcription Starts in Human. Bioinformatics, 22(14):e472–480, 2006.
Cole Trapnell, Brian A Williams, Geo Pertea, Ali Mortazavi, Gordon Kwan, Marijke J van
Baren, Steven L Salzberg, Barbara J Wold, and Lior Pachter. Transcript assembly and
quantification by rna-seq reveals unannotated transcripts and isoform switching during cell
differentiation. Nature Biotech, advance online publication, May 2010. doi:
10.1038/nbt.1621. URL http://dx.doi.org/10.1038/nbt.1621.
A Wolfe, K Singh, Y Zhong, P Drewe, others, G Rätsch, and HG Wendel. Rna g-quadruplexes
cause eif4a-dependent oncogene translation in cancer. Nature, 2014. doi:
10.1038/nature13485.
G Zeller, RM Clark, K Schneeberger, A Bohlen, D Weigel, and G Ratsch. Detecting
polymorphic regions in arabidopsis thaliana with resequencing microarrays. Genome Res, 18
(6):918–929, 2008. ISSN 1088-9051 (Print). doi: 10.1101/gr.070169.107.
A. Zien, G. Rätsch, S. Mika, B. Schölkopf, T. Lengauer, and K.-R. Müller. Engineering Support
Vector Machine Kernels That Recognize Translation Initiation Sites. BioInformatics, 16(9):
799–807, September 2000.

Talk ABRF 2015 (Gunnar Rätsch)

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Talk ABRF 2015 (Gunnar Rätsch)

Similar to Talk ABRF 2015 (Gunnar Rätsch) (20)

Recently uploaded

Recently uploaded (20)

Talk ABRF 2015 (Gunnar Rätsch)